Kristiansen Svein. Maritime Transportation: Safety Management and Risk Analysis
Подождите немного. Документ загружается.
The collision scenario was further modelled as shown in Figure 7.18. This approach
was applied for each of the four segments of the route leading into the harbour. A cargo
spill is dependent on the realization of four different conditions, namely:
1. A ship enters the safe area of the LNG tanker.
2. The two vessels are on collision course.
3. The LNG tanker is the struck vessel.
4. The striking vessel is greater than 1000 tons and has a critical velocity and
collision angle.
Items 1 through 3 can be estimated by the methods outlined in Chapter 6, whereas item 4
can be based on the methods shown in the present chapter.
7. 8 RISK ASSESSMENT FOR A PRODUCTION AND STORAGE TANKER
For any type of operation related to petroleum production on the Norwegian continental
shelf a risk assessment has to be performed. In the final case in this chapter a description
will be given of the analysis of the collision risk for a production and storage tanker (PS).
The assessment (Dahle, 1988, 1992) involves estimation of:
. Probability of collision
. Collision energy
. Loss of safety function
The assessment is summarized in the form of an event tree where the probability of losing
critical, safety functions is computed. The Norwegian petroleum Directorate (NPD) gave
a limiting value of 10
4
on an annual basis.
Figure 7.18. Release due to ship collision branch. (Source: Barlow and Lambert,1979.)
196 CHAPTER 7 DAMA G E ESTIMATION
Case Study
Problem
The following example is a part of an actual analysis called ‘Risk Assessment of
Production and Storage Tankers’. The analysis consists of a traffic modelling part and a
damage modelling part in addition to calculations of the effect of implementation of an
evasive manoeuvre for the PS in a collision situation. In this presentation the damage
modelling part is presented without consideration of an evasive manoeuvre. Based on the
results from the traffic modelling, the damage modelling is perfor med.
Solution
Estimate the collision damage for the production ship (PS). A consequence of a collision
with the PS may be a spill of oil from some of its tank sections. Oil may ignite and cause
further accident escalation by putting shelter areas, escapeways and structure at risk. The
likelihood of these scenarios has to be estimated.
Based on the traffic above, collision probabilities for each traffic segment can be
estimated (Table 7.E1).
Based on the computations above, the total probability (random course situations and
drifting situations) of collision is estimated to 84.5 10
4
.
7.8 RISK ASSESSMENT FOR A PRODUCTION AND STORAGE T ANKER 197
Ta b l e 7. E 1
Vessel type Annual collision
probability P
a
Total time
within zone
per year (hrs)
Conditional probability of impact speed given
a collision situation P
i
Random Drifting Random Drifting
Low High Low High
Visiting Supply 15 10
4
6.2 10
4
55
0.04
0.01 0.1 0
Shuttle tankers 30 10
4
3.1 10
4
110
Nearby 26 10
4
9.9 10
5
1055 0.04 0.01 0.04 0.01
Passing Non-tankers 2.7 10
4
1.6 10
5
108 0.15 0.70 0.04 0.01
Tanker 2.9 10
5
2.1 10
6
11
198 CHAPTER 7 DAMA G E ESTIMATION
Assum pt i ons
For simplicity only two speed ranges are used in the study. Velocities between 3 and
6 knots are considered as low speed while 6 to 15.5 knots is considered as high speed.
The displacement of the PS is 100,000 t. There are large differences in the size of the
shuttle tankers and passing ships. It is assumed that 50% of both the shuttle tankers and
the passing tankers are in ballast condition and the displacement set to 70,000 t while the
other half are fully loaded with displacement of 150,000 t. 95% of the passing non-tankers
are small vessels of 5000 t while the rest of the passing non-tankers are assumed to be
merchant vessels with mean displacement 40,000 t.
The added mass coefficient for the striking ship is 0.1 while for the production ship
it is 0.7. The probability of having a central collision impact is 0.5 and that of a non-
central impact (45
–70
) is 0.25; both situations will lead to hull penetration.
Impact energy
There are two basic collision scenarios. The striking ship may lose prop ulsion power and
drift towards the PS or the striking ship can lose navigational control and continue on a
random course. The two collision types have quite different characteristics with respect to
impact energy and impact geometry. The bow of the PS is heading against the waves. The
wind forces are the dominating load on the drifting vessel. As the angle between the wind
and waves is relatively small, a central impact is not likely to occur. Consequently the
drifting ship may hit the relatively strong bow of the PS and slide along its side. In this case
the impact energy will be low due to the small collision angle and low impact speed. Hence
the damage will most likely be minor and local. The drifting collisions are therefore not
analysed further.
The probability of having a power ed collision is calculated in Table 7.E2.
A ship sailing on a random course will more likely strike the PS in a central impact
given the alternative meeting angles. In addition the impact speed will be higher and there-
fore cause more damage then the drifting vessels. Based on Eq. (7.1) and the assumed added
mass coefficients, the fraction of the kinetic energy transferred in plastic deformation of the
PS’s structure is calculated for all the ships and for both speed categories. The energy
transferred to the production ship (Table 7.E3) is calculated by Eq. (7.2).
Stru ctural dam age
Based on the collision geometry and the necessary crushed hull volume, the probability of
penetrating the hull and cargo tanks is estimated by using Minorsky’s equation (7.3) as
shown in Table 7.E4. It is assumed that in a central impact about 8.5 m
3
crushed material
is necessary to penetrate a cargo tank. Hence the minimum force is 430 MJ. It is further
assumed that the energy to penetrate the hull is 40 MJ. This is a crude assumption as the
value is outside the valid range for Minorsky’s equation. For a non-central impact the
energy necessary is 140 MJ.
It can be seen from Table 7.E4 that high-speed impacts only result in cargo tank
penetration. Based on these results, an event tree is developed as shown below. It is
7.8 RISK ASSESSMENT FOR A PRODUCTION AND STORAGE T ANKER 19 9
Ta b l e 7. E 2
Vessel type Col. 1 Col. 2 Col. 3 Col. 4 Col. 5 Col. 6
Fraction
within
group
Impact
speed
Energy
(MJ),
equation (7.1)
Speed
fraction of
time within
zone
Collision
frequency
(P
a
Col. 1)
Collision
probability
(Col. 5 Col. 4)
Visiting Supply 1 Low 11 0.04 15 10
4
0.6 10
4
High 44 0.01 0.15 10
4
Shuttle tanker 70,000 t 0.5 Low 154 0.02 30 10
4
0.6 10
4
High 616 0.005 0.15 10
4
150,000 t 0.5 Low 330 0.02 0.6 10
4
High 1320 0.005 0.15 10
4
Nearby 1 Low 11 0.04 26 10
4
1.04 10
4
High 44 0.01 0.26 10
4
Passing Small 5000 t 0.95 Low 11 0.12 2.7 10
4
0.32 10
4
High 44 0.56 1.51 10
4
Merchant 40,000 t 0.05 Low 68 0.03 0.08 10
4
High 272 0.14 0.38 10
4
Tanker 70,000 t 0.5 Low 154 0.075 2.9 10
5
0.02 10
4
High 616 0.035 0.01 10
4
Tanker 150,000 t 0.5 Low 330 0.075 0.02 10
4
High 1320 0.035 0.01 10
4
5.9 10
4
200 CHAPTER 7 D AMAG E ESTIMATION
Ta b l e 7. E 3 . Impac t energy transferred to PS in powered collision
Vessel type Col. 1
M
ship
(t)
Col. 2
Impact
speed
Col. 3
Kin. energy
(MJ)
Col. 4
Transferred
energy
Col. 5
Collision energy
(MJ)
Visiting Supply 5000 Low 11 0.97 11
High 44 43
Shuttle tanker Ballast 700,000 Low 154 0.71 109
High 616 437
Loaded 150,000 Low 330 0.53 175
High 1320 700
Nearby 5000 Low 11 0.97 11
High 44 43
Passing Small 5000 Low 11 0.97 11
High 44 43
Merchant 40,000 Low 68 0.81 55
High 272 220
Tanker 70,000 Low 154 0.71 109
High 616 437
Tanker 150,000 Low 330 0.53 171
High 1320 700
7.8 RISK ASSESSMENT FOR A PRODUCTION AND STORAGE T ANKER 201
Ta b l e 7. E 4
Vessel type Impact
speed
No. of penetrations P(no
penetration)
P(penetration
(outside shell))
P(penetration
of cargo tank)
Central
(P ¼0.5)
Non-central
(P ¼0.25)
Visiting Supply Low 0 0 0.6 10
4
00
High 1 0 0.075 10
4
0.075 10
4
0
Shuttle tanker Ballast Low 1 0 0.3 10
4
0.3 10
4
0
High 2 0 0 0.075 10
4
0.075 10
4
Loaded Low 1 1 0.15 10
4
0.45 10
4
0
High 2 0 0 0.075 10
4
0.075 10
4
Nearby Low 0 0 1.04 10
4
00
High 1 0 0.13 10
4
0.13 10
4
0
Passing Small Low 0 0 0.32 10
4
00
High 1 1 0.75 10
4
0.75 10
4
0
Merchant Low 1 0 0.04 10
4
0.04 10
4
0
High 1 1 0.09 10
4
0.29 10
4
0
Tanker Ballast Low 1 0 0.01 10
4
0.01 10
4
0
High 2 0 0 0.005 10
4
0.005 10
4
Loaded Low 1 1 0.005 10
4
0.015 10
4
0
High 2 0 0 0.005 10
4
0.005 10
4
Sum 3.51 10
4
2.22 10
4
0.16 10
4
Fraction 0.60 0.38 0.02
202 CHAPTER 7 DAMA G E ESTIMATION
assumed that there is no loss of safety functions unless spilled oil is ignited. However, if the
spilled oil from at least one cargo tank is ignited, the shelter areas and escapeways are
assumed to be lost. If the engine room and one cargo tank is penetrated the PS will sink,
or in other words there will be loss of all safety functions.
Conclusion
The initial collision probability was 84 10
4
. This value gives a return period of
approximately 100 years. Of these collisions, 88% are due to random navigation. By
eliminating the collisions with insignificant impact energy the risk is reduced by a factor
of 10 to a collision probability of 5.9 10
4
.
The Norwegian Petroleum Directorate (NPD) has a maximum accepted value for
loss of safety functions equal to 10
4
. Neither of the three safety functions has a
probability greater than this target. Hence the PS meets the requirement with respect to
collisions.
7.9 RISK ASSESSMENT F OR A PRODUCTION AND STORAGE T ANKER 203
7.9 R ULES A N D R EGU LATI ON S
A number of regulations influence the design and operation of a ship. These requirements
can roughly be categorized in international, domestic and classification society
regulations. This section gives a brief presentation of the international requirements for
tankers. The conventions relevant for this chapter are:
. The International Convention on Load Lines (ILLC)
. The International Convention for the Safety of Life at Sea (SOLAS)
. The International Convention for the Prevention of Pollution from Ships (MARPOL)
Both ILLC and SOLAS indirectly influence the ability of the vessel to withstand spill
of oil cargo. The ILLC gives standards for the maximum draft and hence limits the cargo
capacity for a given ship size. SOLAS has a broader objective: to ensure the safety of
the ship itself, its crew, passengers, cargo and indirectly the environment. SOLAS covers
a broad range of measures such as subdivisions, stability requirements, construction
principles, safety equipment, fire protection equipment and navigational equipment.
MARPOL gives specific requirements for tankers relating to design, equipment
and procedures for cargo handling. MARPOL is therefore directly addressing pollution
prevention. Some of the key measures are summarized in Table 7.4 (NRC, 1991).
Ta b l e 7. 4 . Elements in MARPOL
Regulation Description
Segregated ballast tanks (SBT) Cargo cannot be carried in ballast tanks
Protective location of SBT SBT are located in ship’s side and bottom
Draft and trim requirements Draft amidships of not less than 2.0 0.02 L and
trim by stern not greater than 0.015 L
Tank size limitations Maximum tank length 10 m–0.2 L depending on location.
Maximum volume may vary up to 22,500 m
3
for side
tanks and 50,000 m
3
for centre tanks
Hypothetical outflow of oil Maximum hypothetical oil outflow after an assumed
damage with specified extents
Subdivision and stability Subdivision is performed to never give a heel larger than
25
and a final water line above any opening as a
consequence of an assumed damage
Crude oil washing (COW) Cargo tanks are cleaned with oil as washing medium
Inert gas system (IGS) Lack of oxygen in cargo tanks
Slop tanks Discharged oil is stored
204 CHAPTER 7 DAMA G E ESTIMATION
REFERENCES
Alexandrov, M., 1970, Probabilistic approach to the effectiveness of ship life-saving systems.
Transactions of SNAME, Vol. 14, 391–416.
Barlow, R. E. and Lambert, H. E., 1979, The Effect of U.S. Coast Guard Rules in Reducing the
Probability of LNG Tanker-Ship Collision in Boston Harbour. May 21, Tera Corporation,
Berkeley, CA.
Dahle, E. A., 1992, Risk Assessment of Production and Storage Tankers. Det Norske Veritas, Høvik.
Dahle, E. A., 1988, Damage control on floating platforms [In Norwegian: Begrensning av konsekvenser
ved skade pa
˚
flytenede konstruksjoner]. Lecture Notes, Division of Marine Systems Design,
Norwegian University of Science and Technology, Trondheim.
Gallagher, P., 1997, A Simple Formula for Approximating the Longitudinal Extent of Bottom Damage
to High Speed Craft Resulting from Hull Penetrations by Submerged Hazards. IMO SLF HSC
Code Correspondence Group, Document 97-02.
IMO, 1993, Explanatory Notes for the SOLAS Regulations on Ship Subdivision and Damage Stability
of Cargo Ship of 100 Meters and Over. Resolution A 684(17), Sales number IMO-871E.
International Maritime Organization, London.
IMO, 1996, Interim Guidelines for the Approval of Alternative Methods of Design and Construction of
Oil Tankers under Regulation 13 F(5) of Annex I of MARPOL 73/78, MARPOL 73/78 – 1994
and 1995 Amendments, Sales number IMO-640E. International Maritime Organization,
London.
Kostilainen, V., 1971, Analysis of Casualties of Tankers in the Baltic, Gulf of Finland and Gulf of
Bothnia in 1960–1969. Helsinki University of Technology, Ship Hydrodynamics Laboratory,
Report No. 5, Otaniemi.
Kostilainen, V. and Hyva
¨
rinen, M., 1976, Ship Casualties in the Baltic, Gulf of Finland and Gulf
of Bothnia in 1971–75. Helsinki University of Technology, Ship Hydrodynamics Laboratory,
Report No. 10, Otaniemi.
McGee, S. P., Troesch, A. and Vlahopoulos, N., 1999, Damage length predictor for high-speed
Craft. Marine Technology , Vol. 36(4), 203–210.
Minorsky, V. U., 1959, An analysis of ship collisions with reference to protection of nuclear power
plants. Journal of Ship Research, Vol. 3(2), 1–4.
NRC, 1991, Tanker Spills Prevention by Design. National Research Council, National Academy
Press, Washington, DC.
Vaughan, H., 1978, Bending and tearing of plate with application to ship-bottom damage. The Naval
Architect, May, 97–99.
Zhang, S., 1999, The mechanics of ship collisions. Ph.D. Thesis, Department of Naval Architecture
and Offshore Engineering, Technical University of Denmark, Lyngby.
205